# ReproHack Brainstorm

ideal: software carpentry template model to make repro-hacks easier to start and do.
- we can use *GitHub templates* and generate event pages from it

opencon london page has a google form to suggest papers to reproduce. Shiny app scrapes the google form data and presents it. Potential hosting problems with Shiny (example: you run out of money on the free account).

Can we quantify the hosting cost?

Using Google Form for data entry is convenient, but tied to a Google Form. Can we do this in a more open way?

Can we use github? EG use github issues as a datastore. Would users have to authenticate to github then accept your app as a connected app?

see also rescience

Can we do it with github pull requests?
- EG, pull request to add a YAML file that describes a paper, Travis CI runs to validate YAML file. when merged Travis CI builds static site.
- [Radovan:] I also support the idea of serving the website for the event as static GH pages (or equivalent) and submit papers as pull requests (using a PR template) which would be reviewed and accepted by the organizers of the event. This way we minimize the toolset and would not require Google Forms etc. Having everything in one place (one repository) would also simplify reuse for other events. Generate from a GH template and that's it: no other accounts and places to set up.

Who is proposing papers to be reproduced?

- Authors nominate papers.
- [Radovan:] Problem can be that authors are difficult to reach via Twitter/mailing lists. In other words the authors who need this may not know about us. How about we allow workshop participants to nominate 3 papers or so each, *however* we would then contact authors and ask them whether they agree their paper to be used in the event. For this we would send authors an "executive summary" of the event so that they see what this is about and what is in it for them (good tips and possibly pull request "for free") without doing tons of reading.

we like the seeding because it allows authors to test their claimed reproducibility, and it lets attendees start with a concrete paper and not spend time finding one.

The admins of the workshops need to have the tools to make it clear what data is being stored and whether privacy controls are in place.